Chat API (API Documentation 🚀)

Chat API for Instabug Backend Engineer Challenge.

Stack

Golang, Gin, Rails, Redis, RabbitMQ, ElasticSearch, Docker

API

Running The API

Clone the project

git clone https://github.com/aminyasser/chat-api.git
cd chat-api

You must have docker and docker-compose installed. The application is all dockerize, you just need to type the following command to run it.

 docker-compose  up --build 

API Documentation

You can check out the postman documentation from Here 🚀 Or import this Postman Collection to try locally.

All requests in the api under port 8000, Example of creating app:

POST    http://0.0.0.0:8000/api/v1/apps

System Architecture

Diagram

Service Structure

I was asked to do specific parts of the application with Golang to handle race conditions, so I made a Golang service for creating chats and messages and running workers that update the chat_count and message_count in apps and chats tables.

The rails service contains creating an app and getting the app endpoints, also containing the search endpoint.

Finally, the apigateway service is a simple go script forwarding the requests to the right service using reverse proxy.

Redis and Race Condition problem

I used Redis INCR to handle the problem of chat_number and message_number, in the problem statement "Numbering of chats in each application starts from 1 and no 2 chats in the same application may have the same number" so I used Redis INCR by the token for chat_count and chat_number:token for message_count.

So because of Redis single-threaded nature, the race condition is avoided here.

There is another way to do that by mutex.lock, but I preferred Redis for doing it.

Message queueing (RabbitMQ)

I used rabbitMQ to handle the amount of requests and avoid running the critical part of creating chats and messages into the database directly.

when the app starts, the queue is declared and the consumer is ready to fetch the publishing messages.

Elastic Search

using elastic search for fast searching queries on messages, I mapped the body, and chat_id.

basic searching with body of the message using standard analyzer and filter with chat_id.

Database

Diagram

Indexes

I used indexes in the frequent queries, I added app_token index on the apps table. to make getting an app query faster, I used it to check if the app exits in the database.

I add also chat_number and message_number in each chat and messages table, with them I search by foreign keys like app_token and chat_id, so the search query will be fast.